Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Caltech101

Benchmarks

Task NameDataset NameSOTA ResultTrend
Image ClassificationCaltech101 Base and New Classes
Base Accuracy98.95
52
Image ClassificationCaltech101 few-shot
Accuracy95.1
32
Multi-view ClusteringCaltech101 20
ACC76.81
30
Fine-grained classificationCaltech101 (base classes)
Accuracy98.23
27
Fine-grained classificationCaltech101 (novel classes)
MCE0.39
15
Few-shot Image ClassificationCaltech101 (test)
Accuracy (1-shot)94.17
15
Image ClassificationN-Caltech101 (test)
Accuracy81.73
13
Fine-grained classificationCaltech101 fine-grained (base)
MCE0.24
12
Image ClassificationCaltech101 New classes
Accuracy94.93
12
Image ClassificationCaltech101 Base classes
Accuracy99.23
12
Image ClassificationCaltech101
Base Accuracy98.97
11
Image ClassificationCalTech101 zero-shot
Clean Accuracy93.6
11
Semi-supervised classificationCaltech101 7
Accuracy97.5
10
ClusteringCALTECH101-7
AMI0.6592
9
Image ClassificationCaltech101 Pathological Non-IID (test)
Accuracy97.02
9
Event-based ClassificationN-Caltech101 (test)
GFLOPs0.7
9
Image Anonymization EvaluationCaltech101
CLIP Score30.94
7
ClassificationCaltech101 20
Accuracy92.48
7
ClusteringCaltech101 20 (test)
Accuracy45.12
7
Few-shot ClassificationCaltech101 16-shot (test)
Accuracy (Sym 0.125)92.07
5
Image ClassificationCaltech101 (unseen)
Accuracy94
4
Image ClassificationCaltech101 (Novel)
Top-1 Acc94.5
4
Image ClassificationCaltech101 (PGD-l∞ attack, ε=4/255) (test)
Robust Acc80.73
4
Image ClassificationCaltech101 PGD-l2 attack, ε=0.5 (test)
Robust Accuracy89.21
4
ClusteringCaltech101 7 (test)
Accuracy88.27
2
Showing 25 of 26 rows